a meta-analysis of maternal prenatal depression and ... · impacts on fetal neurodevelopment....

A Meta-Analysis of Maternal Prenatal Depression and Anxiety on Child Socio-Emotional Development

Author listing: Sheri Madigan1,2, PhD, Hannah Oatley3 MD, Nicole Racine3, PhD, R.M.P. Fearon4 PhD, Lea Schumacher5, Emis Akbari6, PhD., Jessica Cooke1,2, and George M. Tarabulsy7 PhD Affiliations: 1University of Calgary, Calgary, AB, Canada; 2Alberta Children’s Hospital Research Institute, Calgary, AB, Canada; 3Department of Pediatrics, Hospital for Sick Children, Toronto, ON, Canada; 4University College London, London, United Kingdom; 5University of Groningen, Groningen, the Netherlands; 6George Brown College, Toronto, ON, Canada; 7 Laval University, Québec, QC, Canada. Address correspondence to: Sheri Madigan, Department of Psychology, University of Calgary, 2500 University Ave., 2500 University Dr. N.W., Calgary, AB, T2N 1N4, Canada, Phone: (403) 220-5561; Fax: (403) 282-8249; email: [email protected]

Short title: Maternal Prenatal Stress and Child Behavioral Outcomes Financial Disclosures: The authors have no financial relationships relevant to this article to disclose. Funding Source: Research support was provided to the first author by the Alberta Children’s Hospital Foundation and the Canada Research Chairs program. Conflicts of Interest: The authors have no conflicts of interest relevant to this article to disclose. Tables = 1 Figure = 3 Supplemental Tables and Figures = 4 Word Count: 6952

Prenatal Stress and Child Outcomes 1

Running Head: PRENATAL STRESS AND CHILD OUTCOMES

A Meta-Analysis of Maternal Prenatal Depression and Anxiety on

Child Socio-Emotional Development


Abstract

Objective: Observed associations between maternal prenatal stress and children’s socio-

emotional development have varied widely in the literature. The objective of the current study

was to provide a synthesis of studies examining maternal prenatal anxiety and depression and the

socio-emotional development of their children.

Method: Eligible studies through to February 2018 were identified utilizing a comprehensive

search strategy. Included studies examined the association between maternal prenatal depression

or anxiety and the future development of their children’s socio-emotional development (e.g.,

difficult temperament, behavioral dysregulation) up to 18 years later. Two independent coders

extracted all relevant data. Random-effects meta-analyses were used to derive mean effect sizes

and test for potential moderators.

Results: 91 effect sizes from 71 studies met full inclusion criteria for data analysis. The

weighted average effect size for the association between prenatal stress and child socio-

emotional problems was OR = 1.66 (95% CI = 1.54-1.79). Effect sizes were stronger for

depression (OR = 1.79; 95% CI = 1.61-1.99) compared to anxiety (OR = 1.50; 95% CI = 1.36-

1.64). Moderator analyses indicated that effect sizes were stronger when depression was more

severe and when socio-demographic risk was heightened.

Conclusions: Findings suggest that maternal prenatal stress is associated with offspring socio-

emotional development, with the effect size for prenatal depression being more robust than

anxiety. Mitigating stress and mental health difficulties in mothers during pregnancy may be an

effective strategy for reducing offspring behavioral difficulties, especially in groups with social

disadvantage and greater severity of mental health difficulties.

Keywords: Meta-analysis; Prenatal stress; child socio-emotional behavior


Brain development occurs most rapidly during the fetal period1. Consequently, insults to

the intrauterine environment and changes to biochemical signalling pathways can have profound

impacts on fetal neurodevelopment. Maternal prenatal stress, in both human and animal models,

has been recognized as a prenatal programming factor that adversely affects fetal development2-4.

Maternal prenatal stress has also been associated with preterm birth and low birth weight, two

common causes of neurocognitive, as well as socio-emotional and behavioral delays and deficits

in childhood 2.

Maternal prenatal stress has been conceptualized and measured in various ways, but the

most common has been to measure maternal anxiety or depression during pregnancy5. To date, a

vast body of literature has examined the impact of maternal prenatal stress on perinatal and

postnatal health outcomes. Given the variability in results, several meta-analyses have been

performed in an attempt to clarify patterns within this area of research. These meta-analyses have

examined associations between maternal prenatal stress and infant gestational age, birth weight,

and child cognitive outcomes, revealing weak but consistently significant associations 2-4.

However, a meta-analysis examining the sizeable body of research on maternal prenatal anxiety

and depression on children’s socio-emotional development has yet to be conducted.

Socio-emotional development is a general construct that encompasses emotional and

behavioral problems, and difficulties with self-regulation, social information processing and

emotional understanding 6. Socio-emotional competence in early childhood provides a critical

foundation for future academic skills and well-being 7. However, if dysregulation predominates

or maladaptation occurs, delays and deficits in socio-emotional development can lead to mental

health difficulties in adolescence and adulthood 8. By understanding the early determinants of


socio-emotional (mal)adaptation, we can better devise and implement intervention strategies to

support children’s healthy development 9,10.

Several potential moderators of the maternal prenatal stress to child outcome links have

been indicated by past research. One such moderator is the operationalization of constructs that

are indicators of stress, such as anxiety and depression. Indeed, analyzing the differential

influence of maternal prenatal anxiety and depression on child developmental outcomes has been

highlighted as an important area of inquiry for the field11. While there is relative uniformity in

the definition, operationalization, and measurement of maternal depression during pregnancy,

which is most often measured through questionnaires, or less commonly through diagnostic

interviews, there is considerable variability in the operationalization and measurement of

prenatal anxiety. Some studies assess for trait- or state-based anxiety, perceived stress, or

stressful life events, most often via questionnaire measures, and other studies examine

physiological indicators of the stress response, such as cortisol assays. This measurement

variability may be an important contributor to between-study heterogeneity. Another potential

moderator is the timing of exposure to stress. Some studies have suggested that child

development is most impacted by exposure to stress during the second trimester of pregnancy 12,

whereas others argue that stress during the third trimester is most influential13. Finally, greater

clinical severity of depression and/or anxiety, as well as the presence of contextual stressors such

as socio-economic deprivation, may also exacerbate the biological processes involved in the

stress response, and/or increase fetal exposure to prenatal toxins (e.g., drug, alcohol, and tobacco

use), which have implications for child neurodevelopment and behavior 4,14.

The overarching objective of the current study is to provide a quantitative synthesis of the

literature to provide a more precise determination of the magnitude of association between


maternal prenatal stress and children’s socio-emotional development in observational studies. A

meta-analysis with all studies on prenatal stress will be provided, followed by two distinct

analyses on prenatal depression and anxiety. In order to attain a clear understanding of the

methodological factors that may serve to amplify or attenuate observed associations, several

sample and methodological moderators will be examined to determine if they predict between-

study variation.

Methods

Definitional Criteria

The definition of socio-emotional development in the current meta-analysis is guided by

the Center on the Social Emotional Foundations for Early Learning (CSEFEL), who define the

concept as the developing capacity of the child to “experience, regulate, and express emotions in

socially and culturally appropriate ways; and explore the environment and learn—all in the

context of family, community, and culture15 (pg. 2)”. This definition includes social and

emotional competence (e.g., understanding and selecting appropriate social or emotional

responses), temperament (e.g., fussiness; negative affectivity), behavioral problems (e.g.,

internalizing and externalizing problems), and crying or colic 16.

Search Strategy

This meta-analysis was based on recommendations by PRISMA for reporting systematic

reviews 17. Searches were conducted up to February 2018 in MEDLINE, EMBASE, PsycINFO,

and Cochrane databases for published and unpublished studies. Relevant database specific

subject headings and text word fields were searched (see Table S1, available online).

Synonymous terms were combined with the Boolean “OR”, and then these concepts were

combined using a Boolean “AND”. Truncation symbols were used in searches when appropriate


to capture variant endings and spellings of search terms. No date restrictions were applied, but

the search was limited to English language articles.

Study Inclusion and Exclusion Criteria

To identify studies meeting inclusion criteria, titles and abstracts identified in the search

strategy were reviewed. Inclusion criteria for the meta-analysis were: (1) maternal depression

and/or anxiety was measured in pregnancy; (2) offspring outcomes were collected prior to the

age of 18y; (3) the study statistic could be transformed into an effect size; and (4) the full-text

article was available and written in English.

A protocol was developed so that each sample of participants was only represented once

in the meta-analysis. First, if a study presented more than one predictor (e.g., trait anxiety and

pregnancy fears) or outcome (e.g., temperament and behavior problems), a mean effect size

across measures was calculated and entered into the meta-analysis. One exception to this rule

was if a study presented separate effect sizes for the association between prenatal depression and

child socioemotional outcomes and prenatal anxiety and child outcomes, where both were

included in the analyses. Second, if more than one time point of maternal prenatal stress or child

outcomes were provided, effect sizes were pooled across time points. The average prenatal time

point or child age of those pooled effect sizes was used in moderator analyses. Third, if multiple

publications emerged from a dataset, we selected the publication with the largest sample size and

most comprehensive data extraction information.

Data Extraction


Studies meeting inclusion criteria were coded using a standard data extraction form.

Potential moderators included: type of child outcome (colic/crying, temperament, behavior

problems), method of assessing child outcome (questionnaire, structured interview, or

observation), socio-demographic risk (e.g., low income, low education, single parent), maternal

age at pregnancy, gestational age at the time of the prenatal stress measurement, child age at the

outcome assessment, percent of children who are male, the type of prenatal anxiety measure

(state, trait/pregnancy-stress, cortisol, life events, or mixed measurements), postnatal control for

depression/anxiety (yes/no), clinical risk (i.e., syndromal-level of depressive symptoms versus

not), and study quality score. Approximately 15% of included studies were double coded for the

purpose of establishing reliability of data extraction. For categorical moderators, the percent

agreement for all moderators was 100%, for continuous moderators the inter-coder agreement

ranged from .79 to .99, and for effect size extraction it was .99. Discrepancies were resolved by

consensus coding.

Additionally, an assessment of study quality was conducted based on a 15-point quality

assessment tool adapted from the National Institutes of Health Quality Assessment Tool for

Observational Cohort and Cross-Sectional Studies (2014)18 (see Table S2, available online). This

assessment rating tool evaluates elements of study quality endorsed by the Cochrane

collaboration19. Data extraction was conducted by a primary coder, and a proportion of the

studies (15%) were verified by a second coder. The inter-coder agreement was .77, and

discrepancies were resolved by consensus coding.

Data Analysis


We used a two-step approach to examining mean effect sizes. First, as several studies had

assessments of both anxiety and depression, we initially conducted a multivariate test of

dependent effect sizes using the robust variance approach methods developed by Fisher and

Tipton20 and the R package robumeta. Second, for the univariate sub-analyses examining the

depression to child outcome, as well as the anxiety to child outcome link, we conducted

univariate tests of non-overlapping studies using Comprehensive Meta-Analysis (CMA, version

3.0)21 software, to test for overall effect sizes, publication bias, and potential moderators.

Effect sizes in individual studies were transformed into Odds Ratios (OR), and 95%

confidence intervals (95% CI) around the mean point estimate are provided. Effect size

calculations were based on random effects modelling based on the assumption that each study

has its own population parameters. To assess for heterogeneity of effect sizes, the Q and I2

statistics were computed. A significant Q statistic suggests that study variability in effect size

estimates is greater than sampling error and that moderators should be explored. The I2 statistic

examines the rate of variability across studies due to heterogeneity rather than chance, with

values of 50% and 75% or above suggesting moderate to high heterogeneity, respectively22. To

examine whether moderators could explain variability across studies, categorical and continuous

moderators were conducted using Q statistics and meta-regressions21, respectively. Publication

bias was examined using Duval and Tweedie’s trim and fill method23, as well as Egger’s Test. If

publication bias was indicated, the trim and fill procedure was used to correct for publication

bias 23,24.

Results

Studies Selected


The PRISMA flow diagram17 detailing the search strategy and resulting outcomes can be

found in Figure 1. Our electronic search of four databases yielded 8,814 articles after duplicates

were removed. Upon review of the titles and abstracts, 518 articles were identified as potentially

meeting study inclusion criteria. After further review of full text articles, 73 studies met full

inclusion criteria.

Study Characteristics and Quality

Study characteristics are reported in Table 1. The sample sizes of studies ranged from 24

to 8,328 parent–child dyads. The average prenatal stress time point was 27.74 weeks gestation

(range = 7.5-37 weeks). At the outcome assessment, child age averaged 30.59 months (range =

range .2 months to 17 years). Thirty-one studies were from North America (42.47%), 29 from

Europe (39.73%), eight from Australia (10.96%), 4 from Asia (5.48%), and one from Africa

(1.41%). For study quality, the mean score across all studies was 8.36 (SD=2.1; range 4-14; see

Table S2, available online).

Multivariate Testing of Dependent Effect Sizes: Maternal Prenatal Stress and Children’s

Socio-Emotional Development

Studies with an effect size value larger or smaller than +3 standard deviations from the

mean were considered outliers. Two studies were identified and removed from subsequent

analyses.

Multivariate Testing of Dependent Effect Sizes. Over all the anxiety and depression

outcomes (91 effect sizes from 71 studies), the weighted average effect size was OR = 1.66

(natural log Odds Ratio = .51, SE = .038, p <.001, OR 95% CI = [1.54,1.79]. I2 was estimated to


be 51%. A test of the difference in effect sizes between anxiety and depression outcomes showed

that depression was associated with a significantly larger effect than anxiety (difference in

natural log odds ratio = .18, SE = .065, p = .009; 95% CI = [.046, .311]). The average effect size

for the depression studies was OR = 1.79 (95% CI = [1.61, 1.99]), whereas for anxiety it was OR

= 1.50, 95% CI = [1.36, 1.64]). Sensitivity analyses showed that different assumed values for

the within-study covariance had a trivial impact on the model estimates (test values ranged from

0 to 1 in .20 intervals). Given the statistically significant difference in effect sizes for the

depression-child outcome link and anxiety-child outcome link, subsequent analyses were

conducted separately for each construct.

Univariate Analyses: Maternal Prenatal Depression and Children’s Socio-Emotional

Development

A total of 50 non-overlapping studies (33,211 mother-child dyads) were available to

estimate the mean effect size for the association between prenatal depression and socio-

emotional development. A random-effects meta-analysis produced a significant combined effect

size of OR = 1.76 (CI: 1.60-1.94). Figure 2 depicts the forest plot. Egger’s test suggested

publication bias (p<.05); and the funnel plot was asymmetric (see Figure S1, available online).

Using the trim-and-fill procedure with 8 studies imputed, the adjusted effect size was OR = 1.63

(CI: 1.47-1.81).

Statistically significant heterogeneity between the studies was found (Q = 86.4, p < .001;

I2 = 43.28). Effect sizes were stronger in studies examining depression diagnostically (k=13, OR

= 2.26; CI: 1.86-2.78) versus depressive symptoms (k=37, OR = 1.60; CI: 1.46-1.75). The

proportion of between-study variance explained by this moderator was R2 = .47. The association


between prenatal depression and socio-emotional development was stronger in studies

characterised as being socio-demographically at risk (k=10, OR = 2.24; CI: 1.73-2.91) compared

to studies without such risk (k=40, OR = 1.66; CI: 1.51-1.84). The proportion of variance

explained by this moderator was R2 = .15. The remaining moderators tested did not reliably

explain between-study heterogeneity, including type and method of assessing child behavior,

pregnancy time point, maternal age, and child age and gender (see Table S4, available online).

Univariate Analyses: Maternal Prenatal Anxiety and Children’s Socio-Emotional

Development

With the two outliers removed, a total of 41 non-overlapping studies (17,799 mother-

child dyads) were included in the meta-analysis on prenatal anxiety and socio-emotional

development. In CMA, the point estimate for this association was OR = 1.47 (CI: 1.36-1.57).

Figure 3 depicts the forest plot. Egger’s test for publication bias was significant (p<.01), and the

funnel plot revealed asymmetry (see Figure S2, available online). Using the trim-and-fill

procedure, with 12 studies imputed, the adjusted effect size was OR = 1.33 (CI: 1.23-1.44).

Statistically significant heterogeneity between the studies was found (Q = 56.63, p < .05, I2 =

29.36). Moderator analyses were explored, including type of prenatal stress measure, type of

child outcome, postnatal controls, method of assessing child behavior, pregnancy time point,

maternal age, and child age and gender (Table S5, available online), however none of these

moderators explained between-study heterogeneity.

Discussion

Results from this meta-analysis are consistent with the view that prenatal depression and

prenatal anxiety have an adverse effect on children’s socio-emotional and behavioral


functioning. Specifically, for mothers experiencing prenatal depression and anxiety, the odds of

having children with behavioral difficulties were almost 1.5-2 times greater than those not

experiencing prenatal depression or anxiety. Although this is the first meta-analysis to rigorously

evaluate the literature on maternal prenatal stress and socio-emotional development, our findings

build on several studies examining associations between maternal prenatal stress and other child

outcomes, including birth weight, gestational age at delivery, and cognitive development 2-4.

Amassing evidence from current and past meta-analyses, maternal prenatal stress, broadly

defined, clearly has robust associations with, and may have deleterious effects on, children’s

developmental health.

The specific mechanisms through which maternal prenatal stress impact brain

development and later child outcomes are not well understood. The most widely investigated

hypothesis involves programming via the maternal hypothalamic-pituitary-adrenal (HPA) axis,

which produces elevated levels of cortisol under stress, which in turn may subsequently disrupt

fetal brain development. However, research has shown that the maternal HPA axis is less

responsive to stress during pregnancy and that the placenta acts as a shield offering protection

against maternal hormones, calling into question the extent to which maternal stress hormones

reach the fetus 25. Thus, a more recent theory has hypothesized that the placenta may play a key

regulatory role in the mechanisms underlying fetal programming 26. During pregnancy the

placenta produces the 11B-HSD2 enzyme, which works to transform active cortisol into inactive

cortisone 27. Both human and animal models of maternal prenatal stress have shown a reduction

in the expression and activity of the 11B-HSD2 placental enzyme under conditions of maternal

prenatal stress 28,29, resulting in increased fetal exposure to maternal hormones. Furthermore,

maternal stress, including prenatal depression and anxiety, has been associated with an increased


risk of preeclampsia in later pregnancy 30. Mothers with preeclampsia show higher levels of

placental CRH (corticotropin-releasing hormone) 31 and reduced blood serum levels of Placental

Growth Factor (PlGF) 32. Exposure to both these maternal hormones prenatally may negatively

impact the infant’s brain and HPA axis development, which have strong associations with socio-

emotional and behavioral functioning across the lifespan 33.

There are other mechanisms whereby maternal prenatal stress impacts neural

development. As previously discussed, mothers who experience stress are more likely to have

preterm and low birth weight babies. Maternal prenatal stress may also affect uterine blood

supply and nutrient transport, thus contributing to fetal growth restriction. Maternal prenatal

stress has been shown to be associated with reduced blood flow in uterine or umbilical arteries 27,

as well as changes in fetal cerebral circulation 34. A more direct effect of gestational stress on

placental nutrient transport has been documented in rodents 35. These restrictions in turn may

contribute to compromised neural development and poorer behavioral outcomes. Finally, it is

important to note that most studies cannot rule out genetic confounding as an explanation,

although one study using an in-vitro-fertilization design found some evidence for environmental

transmission of maternal stress to child conduct problems36.

We found that effect sizes were stronger in studies that examined depression at the

syndromal-level. Not only may clinical depression harm fetal development through the stress

mechanisms described above, but women with clinical depression are also less likely to seek out

prenatal care compared to those without 37, and have been reported to have poorer overall

physical health (e.g., eating, sleeping, and exercise behaviors), as well as psychological health, in

the prenatal period 26. These factors in turn may lead to alterations in biophysiological

functioning and fetal development38. The prenatal health risk behaviours associated with greater


clinical depression severity, may also explain the stronger effect sizes for the association

between prenatal depression and child socio-emotional outcomes compared to that for maternal

prenatal anxiety. Prenatal depression may also be more strongly associated with other prenatal

risk factors such as smoking, alcohol use, socio-economic deprivation, and adverse life

circumstances14,39.

There are several other possible explanations for the differential association between

prenatal depression versus anxiety and child socio-emotional outcomes. Prenatal depression has

been associated with dysregulation of the HPA axis as well as other biological and physiological

changes, which in turn could affect fetal biological and physiological development. In addition,

postnatal depression has been linked to maternal emotional unavailability and

unresponsiveness40, which are linked to delays in development. This draws attention to the

critical role of the postnatal environment for the developing child; when depression is maintained

or exacerbated in the postnatal period, it may jeopardize child socio-emotional development

further. Finally, maternal reports are often used to assess problematic child outcomes and

maternal depression can distort perceptions, leading to an overestimation of child behavioral

problems 41. This methodological limitation could influence the magnitude of associations and

should be an important consideration in future research.

Moderator analyses revealed that the association between prenatal depression and socio-

emotional development was also heightened in families characterized as experiencing socio-

economic deprivation. This finding is consistent with a meta-analysis on prenatal stress and

perinatal risks 4. It is well documented that there are large socio-economic disparities in prenatal

and perinatal health. In comparison to mothers from middle to high income groups, mothers from

lower socio-economic strata are more likely to have inadequate prenatal care, higher rates of


prenatal (e.g., intrauterine growth restriction) and perinatal risks and complications (e.g., low

birthweight; preterm birth), leading to increased risk of infant morbidity and mortality 14,39.

Poverty is also associated with a clustering of child-related risks such as food insecurity, chaotic

living arrangements, community violence, and stressful events, which have downstream

consequences for child development 14. The developmental resources that are known to enhance

child development may also not be attainable for families struggling financially, creating further

disparities in children’s socio-emotional development.

We did not find that the type of anxiety measure examined (e.g., state, trait, cortisol) in

pregnancy differentially predicted child socio-emotional development. Moreover, the timing of

the assessment of prenatal anxiety in pregnancy was not found to moderate the association

between prenatal stress and socio-emotional development. The majority of past studies are

limited by the fact that prenatal stress is typically only examined at one time point 4. This may

create a problem in distinguishing between timing of stress and chronicity of stress. Timing of

stress is not analogous with chronicity of stress, and those who report on stress later in their

pregnancy may have experienced more chronic and sustained levels of stress, which could be

more detrimental to the fetus. A ripe avenue of future longitudinal research is the examination of

maternal stress at multiple, discrete time points to provide clarity in this regard.

There are two additional hypotheses worthy of mention that could not be addressed in the

current meta-analyses. First, this meta-analysis was unable to examine the potential impact of

genetic predisposition to behavioral difficulties for children of mothers who demonstrate higher

levels of depression and anxiety. Second, the interaction between maternal prenatal stress and the

quality of the postnatal environment could not be examined herein. Child development is

critically influenced by the postnatal environment. It has been demonstrated that infants born to


mothers experiencing maternal prenatal stress can be buffered from negative child outcomes if

they experience postnatal environments enriched with sensitive and response caregiving behavior

42, as well as secure attachment42. Thus, the quality of parenting, attachment, and the provision of

social supports may attenuate associations between prenatal stress and child socio-emotional

development.

Clinical Implications

An important caveat prior to a discussion of the clinical implications of our findings is

that meta-analyses are correlational in nature. Causality between maternal prenatal stress and

adverse child socio-emotional outcomes cannot be established. Nevertheless, based on the

collective body of research on prenatal stress and deleterious pregnancy, perinatal, and child

outcomes, there is a growing awareness of the importance of early intervention services and

educational campaigns to mitigate stress and foster healthy fetal neurodevelopment and child

outcomes. It is likely that investing in maternal prenatal stress interventions will have broad

reaching beneficial effects on the health of mothers, children and their relationship.

Our findings suggest further that the application of preventive interventions in pregnancy

may be particularly appropriate for mothers presenting with depression and/or those burdened by

demographic risk. However, the current knowledge base for the efficacy of mental health

interventions aimed directly at reducing stress and mental health in pregnancy is limited and

underdeveloped 43-45. Research must be conducted on maternal stress interventions during

pregnancy and longitudinal follow-up over childhood. Future research should also examine the

comparative benefits that different stress reduction strategies, including cognitive-behavioral

therapy, interpersonal therapy, and mindfulness approaches, have on child development 46,47.


Moreover, the impact of interventions on women who experience co-morbid depression and

anxiety is important, as higher cortisol levels have been found in pregnant women with comorbid

anxiety and depression, putting these women’s offspring at particular risk of poor developmental

outcomes 48.

Limitations and Future Directions

Meta-analytic results from the current study must be interpreted in the context of several

limitations. First, publication bias was detected, suggesting that non-significant results were less

likely to be published (file-drawer problem) 49; however adjusted effect sizes were derived to

account for publication bias. Although some studies controlled for ongoing maternal stress and

psychopathology in the post-partum period, the majority did not. Research on the role of

postnatal paternal psychopathology and children’s socio-emotional development, in interaction

with, or over and above the influence of, prenatal or postnatal maternal stress is sparse.

Furthermore, data from fathers can also afford an important test of confounders, as prenatal

effects should only be observed in mothers if they are due to intra-uterine mechanisms. Further,

many of the studies that have examined the impact of maternal prenatal stress have been

conducted in low-risk population samples. This is especially true for prenatal anxiety. It is

possible that biological and social mechanisms influencing the association between maternal

prenatal stress and child development outcomes could vary based on biological and social

vulnerability. Finally, underlying depression and anxiety is a common psychopathologic

dimension50, with shared biological and psychological etiologies51; however, the co-morbidity of

depression and anxiety on child socio-emotional outcomes was not examined in the current

meta-analysis.


All of these limitations warrant future investigation in individual studies with superior

methodological quality to enhance understanding of pathways of transmission, and to elucidate

moderating factors that can foster resilience in children exposed to prenatal stress. There is

strong evidence from animal models 52 that child development outcomes for fetuses exposed to

stress in utero improve after a combination of social and physical enrichment, but human studies

are needed to support this body of work. Previous research has demonstrated that attachment

status moderates the association between maternal prenatal stress exposure and child cognitive

development, whereby the association is diminished for children with a secure parent-child

attachment 42; however, this research should be extended to child socio-emotional outcomes, to

further inform intervention research.

Conclusions

This research adds support to the increasing body of literature suggesting that prenatal

depression and anxiety are potential fetal programming factors, affecting biological, cognitive,

and behavioral development in offspring. While overall effect sizes are not large, they suggest,

as other research has, a consistent association between maternal prenatal stress, and child socio-

emotional development. Extant results from the literature point to the potential importance of the

intrauterine environment in setting certain conditions for child development. Such conditions

may interact with genetic, epigenetic and postnatal factors to influence a range of social,

emotional and behavioral outcomes. Our results suggest an increased risk to children exposed to

prenatal depression, as well as to families who are socially disadvantaged. Investing in

disadvantaged families with young children has a high return on investment53, improving

outcomes for parents, children and their families and avoiding later, higher-cost interventions.


References

1. Joseph R. Fetal Brain Behavior and Cognitive Development. Developmental Review.

2000;20(1):81-98.

2. Grote NK, Bridge JA, Gavin AR, Melville JL, Iyengar S, Katon WJ. A meta-analysis of

depression during pregnancy and the risk of preterm birth, low birth weight, and

intrauterine growth restriction. Archives of General Psychiatry. 2010;67(10):1012-1024.

3. Tarabulsy GM, Pearson J, Vaillancourt-Morel M-P, et al. Meta-Analytic Findings of the

Relation Between Maternal Prenatal Stress and Anxiety and Child Cognitive Outcome.

Journal of Developmental & Behavioral Pediatrics. 2014;35(1).

4. Bussières E-L, Tarabulsy GM, Pearson J, Tessier R, Forest J-C, Giguère Y. Maternal

prenatal stress and infant birth weight and gestational age: A meta-analysis of prospective

studies. Developmental Review. 2015;36:179-199.

5. Glover V. Annual Research Review: Prenatal stress and the origins of psychopathology:

an evolutionary perspective. Journal of Child Psychology and Psychiatry.

2011;52(4):356-367.

6. Campbell SB, Denham SA, Howarth GZ, et al. Commentary on the review of measures

of early childhood social and emotional development: Conceptualization, critique, and

recommendations. Journal of Applied Developmental Psychology. 2016;45(Supplement

C):19-41.

7. Gyllenberg D, Sourander A, Niemelä S, et al. Childhood predictors of later psychiatric

hospital treatment: findings from the Finnish 1981 birth cohort study. European Child &

Adolescent Psychiatry. 2010;19(11):823-833.


8. Bornstein MH, Hahn C-S, Haynes OM. Social competence, externalizing, and

internalizing behavioral adjustment from early childhood through early adolescence:

Developmental cascades. 2010;22(4).

9. Bale TL, Baram TZ, Brown AS, et al. Early life programming and neurodevelopmental

disorders. Biol Psychiatry. 2010;68.

10. Lewis AJ, Galbally M, Gannon T, Symeonides C. Early life programming as a target for

prevention of child and adolescent mental disorders. BMC Medicine. 2014;12(1):33.

11. Glover V. Maternal depression, anxiety and stress during pregnancy and child outcome;

what needs to be done. Best Practice & Research Clinical Obstetrics & Gynaecology.

2014;28(1):25-35.

12. O’Donnell K, O’Connor TG, Glover V. Prenatal Stress and Neurodevelopment of the

Child: Focus on the HPA Axis and Role of the Placenta. Developmental Neuroscience.

2009;31(4):285-292.

13. Davis EP, Glynn LM, Dunkel Schetter C, Hobel C, Chicz-Demet A, Sandman CA.

Corticotropin-releasing hormone during pregnancy is associated with infant

temperament. Dev Neurosci. 2005;27(5):299-305.

14. Jensen SKG, Berens AE, Nelson CA. Effects of poverty on interacting biological systems

underlying child development. The Lancet Child & Adolescent Health. 2017;1(3):225-

239.

15. Yates T, Ostrosky, M. M., Cheatham, G. A., Fettig, A., Shaffer, L., & Santos, R. M. .

Research synthesis on screening and assessing social-emotional competence. .

2008;2017(Nov 1):1-19. http://csefel/.

vanderbilt.edu/documents/rs_screening_assessment.pdf.

http://csefel/


16. Halle TG, Darling-Churchill KE. Review of measures of social and emotional

development. Journal of Applied Developmental Psychology. 2016;45(Supplement C):8-

18.

17. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic

reviews and meta-analyses: The PRISMA statement. Annals of Internal Medicine.

2009;151(4):264-269.

18. Lung NH, Institute B. Quality assessment tool for observational cohort and cross-

sectional studies. Available onlin e: http://www/ nhlbi nih gov/health-pro/guidelines/in-

develop/cardiovascular-risk-reduction/tools/cohort (accessed on 30 October 2016).

2014.

19. Downes MJ, Brennan ML, Williams HC, Dean RS. Development of a critical appraisal

tool to assess the quality of cross-sectional studies (AXIS). BMJ Open. 2016;6(12).

20. Fisher Z, Tipton E. Robumeta: An R-package for robust variance estimation in meta-

analysis. arXiv preprint arXiv:150302220. 2015.

21. Borenstein M, Hedges LV, Higgins JP, Rothstein HR. Comprehensive Meta-Analysis

Version 3. Englewood, NJ: Biostat; 2014.

22. Higgins JPT, Thompson SG, Deeks JG, Altman DG. Measuring inconsistency in meta-

analyses. BMJ: British Medical Journal. 2003;327(7414):557.

23. Duval S, Tweedie R. Trim and fill: A simple funnel-plot–based method of testing and

adjusting for publication bias in meta-analysis. Biometrics. 2000;56(2):455-463.

24. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple,

graphical test. British Medical Journal. 1997;315(7109):629-634.

http://www/


25. Burton GJ, Barker DJP, Moffett A, Thornburg K. The placenta and human

developmental programming. Cambridge University Press; 2010.

26. Beijers R, Buitelaar JK, de Weerth C. Mechanisms underlying the effects of prenatal

psychosocial stress on child outcomes: beyond the HPA axis. European Child &

Adolescent Psychiatry. 2014;23(10):943-956.

27. Wyrwoll CS, Holmes MC, Seckl JR. Review. Frontiers in Neuroendocrinology.

2011;32(3):265-286.

28. Welberg LAM, Thrivikraman KV, Plotsky PM. Chronic maternal stress inhibits the

capacity to up-regulate placental 11β-hydroxysteroid dehydrogenase type 2 activity.

Journal of Endocrinology. 2005;186(3):R7-R12.

29. O'Donnell K, O'Connor T, Glover V. Prenatal stress and neurodevelopment of the child:

focus on the HPA axis and role of the placenta. Dev Neurosci. 2009;31.

30. Landsbergis PA, Hatch MC. Psychosocial Work Stress and Pregnancy-Induced

Hypertension. Epidemiology. 1996;7(4).

31. Hobel CJ, Dunkel-Schetter C, Roesch SC, Castro LC, Arora CP. Maternal plasma

corticotropin-releasing hormone associated with stress at 20 weeks’ gestation in

pregnancies ending in preterm delivery. American Journal of Obstetrics and Gynecology.

1999;180(1, Supplement 2):S257-S263.

32. Torry DS, Wang H-S, Wang T-H, Caudle MR, Torry RJ. Preeclampsia is associated with

reduced serum levels of placenta growth factor. American Journal of Obstetrics and

Gynecology. 1998;179(6, Part 1):1539-1544.


33. Hostinar CE, Gunnar MR. Future Directions in the Study of Social Relationships as

Regulators of the HPA Axis Across Development. J Clin Child Adolesc Psychol.

2013;42(4):564-575.

34. Sjöström K, Valentin L, Thelin T, Maršál K. Maternal anxiety in late pregnancy and fetal

hemodynamics. European Journal of Obstetrics & Gynecology and Reproductive

Biology. 1997;74(2):149-155.

35. Mairesse J, Lesage J, Breton C, et al. Maternal stress alters endocrine function of the

feto-placental unit in rats. American Journal of Physiology - Endocrinology And

Metabolism. 2007;292(6):E1526.

36. Rice F, Harold GT, Boivin J, van den Bree M, Hay DF, Thapar A. The links between

prenatal stress and offspring development and psychopathology: disentangling

environmental and inherited influences.

37. Heaman MI, Moffatt M, Elliott L, et al. Barriers, motivators and facilitators related to

prenatal care utilization among inner-city women in Winnipeg, Canada: a case–control

study. BMC Pregnancy and Childbirth. 2014;14(1):227.

38. Buss C, Entringer S, Moog NK, et al. Intergenerational Transmission of Maternal

Childhood Maltreatment Exposure: Implications for Fetal Brain Development. Journal of

the American Academy of Child & Adolescent Psychiatry. 2017;56(5):373-382.

39. Kramer MS, Séguin L, Lydon J, Goulet L. Socio-economic disparities in pregnancy

outcome: why do the poor fare so poorly? Paediatric and Perinatal Epidemiology.

2000;14(3):194-210.

40. Field T. Maternal Depression Effects on Infants and Early Interventions. Preventive

Medicine. 1998;27(2):200-203.


41. Fergusson DM, Lynskey MT, Horwood LJ. The effect of maternal depression on

maternal ratings of child behavior. Journal of Abnormal Child Psychology.

1993;21(3):245-269.

42. Bergman K, Sarkar P, Glover V, O'Connor TG. Maternal prenatal cortisol and infant

cognitive development: moderation by infant-mother attachment. Biol Psychiatry.

2010;67.

43. Dennis CL. Psychosocial and psychological interventions for prevention of postnatal

depression: systematic review. BMJ. 2005;331.

44. Fontein-Kuipers YJ, Nieuwenhuijze MJ, Ausems M, Budé L, de Vries R. Antenatal

interventions to reduce maternal distress: a systematic review and meta-analysis of

randomised trials. Bjog. 2014;121(4):389-397.

45. O'Connor TG, Monk C, Fitelson EM. Practitioner Review: Maternal mood in pregnancy

and child development – implications for child psychology and psychiatry. Journal of

Child Psychology and Psychiatry. 2014;55(2):99-111.

46. Wilkinson EL, O’Mahen HA, Fearon P, et al. Adapting and testing a brief intervention to

reduce maternal anxiety during pregnancy (ACORN): study protocol for a randomised

controlled trial. Trials. 2016;17(1):156.

47. Lenze SN, Potts MA. Brief Interpersonal Psychotherapy for depression during pregnancy

in a low-income population: A randomized controlled trial. Journal of Affective

Disorders. 2017;210:151-157.

48. Evans LM, Myers MM, Monk C. Pregnant women’s cortisol is elevated with anxiety and

depression — but only when comorbid. Archives of Women's Mental Health.

2008;11(3):239.


49. Rosenthal R. The file drawer problem and tolerance for null results. Psychological

Bulletin. 1979;86:638-641.

50. Steer RA, Clark DA, Beck AT, Ranieri WF. Common and specific dimensions of self-

reported anxiety and depression: the BDI-II versus the BDI-IA. Behav Res Ther.

1999;37(2):183-190.

51. Gutteling BM, de Weerth C, Willemsen-Swinkels SH, et al. The effects of prenatal stress

on temperament and problem behavior of 27-month-old toddlers. Eur Child Adolesc

Psychiatry. 2005;14(1):41-51.

52. McCreary JK, Metz GAS. Environmental enrichment as an intervention for adverse

health outcomes of prenatal stress. Environmental Epigenetics. 2016;2(3):dvw013.

53. Heckman JJ. Invest in early childhood development: Reduce deficits, strengthen the

economy. The Heckman Equation. 2010.

Figure 1 PRISMA flow diagram detailing the search strategy.

Records identified through database searching

(n = 12643)

Scre

enin

g In

clud

ed

Elig

ibili

ty

Iden

tific

atio

n

Additional records identified through other sources

(n = 48)

Records after duplicates removed (n = 8814)

Records screened (n = 8814)

Records excluded (n = 8,296)

Full-text articles assessed for eligibility

(n =518)

Full-text articles excluded, lack of relevant data,

overlapping samples, failed to meet inclusion criteria

(n = 445 )

Studies meeting inclusion criteria (n =73)

Figure 2. Forest plot of the overall mean effect size, as well as the effect size for each study included in the meta-analysis on prenatal depression and child socio-emotional development. Legend: Observed effect sizes (OR) and 95% confidence intervals are indicated for each study included in the meta-analysis.

Figure 3. Forest plot of the overall mean effect size, as well as the effect size for each study included in the meta-analysis on prenatal anxiety and child socio-emotional development. Legend: Observed effect sizes (OR) and 95% confidence intervals are indicated for each study included in the meta-analysis.

Table S1: Summary of Search Strategy Database: PsycINFO <1806 to February week 1 2018> Search Strategy: --------------------------------------------------------------------------------

1. pregnan*/ or trimester/ 2. anxiet*/ or anxious*/ 3. depression*/ or depressive/ 4. stress/ 5. or/2-4 6. infant/ or infancy/ or newborn*/ or baby/ or babies/ or child*/teen*/or adolesc*/or

youth*/ 7. behavio*/develop*/temperament/ 8. internali*/externali*/ 9. emotion*/aggressi*/ 10. or/7-9 11. 1 and 5 12. 6 and 10 13. limit 12 to (childhood <birth to age 12 yrs> or adolescence <age 13 to 17 yrs>) 14. 11 and 13

--------------------------------------------------------------------------------

Table S2 Study Quality Criteria Evaluation1

1. Defined Sample Study has a defined eligibility and exclusion criteria for their sample; and time period (dates) and location (s) of recruitment and assessment.

0 = No 1 = Yes

2. Representative Sample Is the sample representative of a defined population? (i.e. was everyone included who should be and is this sample generalizable) E.g. only selecting mothers of children with disabilities = 0. 1 = Cohorts recruited from the general population or from multi-site studies and/or large databases. 0 = Single site clinical studies.

0 = No 1 = Yes

3. Adequate Sample Size Power calculation provided 0 = No 1 = Yes

4. Participation/Attrition

Does the study meet satisfactory participation/attrition rates? 0= <60% participation; >40% attrition or not specified 1= 60-79% participation; 21-39% attrition 2= >80% participation, <20% attrition

0 = Not-acceptable 1= Marginally acceptable 2 = Acceptable

5. Missing data Does the study mention missing data and account for how they were treated in the analysis? Studies that remove incomplete data from the outset and do not include it in the total N are considered meeting the criteria for addressing missing data

0= No 1= Yes

6. Valid Instrument (Stress) Does the study use a validated instrument for the assessment of maternal prenatal stress? 0 = Non-validated (made up by researcher) 1= validated measure (e.g. Perceived stress scale, parenting stress index, STAI, BDI, EPDS)

0 = Non-validated 1 = Validated

7. Valid Instrument (Child Outcome)

Does the study use a validated instrument for the assessment of child outcome?

0 = Non-validated 1 = Validated

0 = Non-validated (made up by researcher) 1= validated measure (BSID, IBR, CBCL , BASC, IBQ-R, ASQ)

8. Subjective vs. objective measures (Stress)

Does the study use different reporters or multiple-methods to measure maternal stress? Objective measure = cortisol, substantiated life stress, diagnosis or physician evaluation. Multiple methods = self-report and biological data/substantiated life stress, diagnosis.

0 = Self-report 1 = Objective measure 2 = Multiple methods

9. Subjective vs. objective measures (Child Outcome)

Does the study use different reporters or methods to measure child outcome?

0 = Maternal report 1 = Objective measure 2 = Multiple methods (e.g. two reporters, maternal report and observational data)

10. Confounding Variables Were confounding variables taken into account in the analysis? E.g. in modeling or regression did they control or adjust for confounding factors.

0 = No 1 = Yes

11. Demographic Information

Does the study report complete demographic data for parents and children included in the study? 0 = No demographic information specified 1 = data for only child or parent 2= demographic data for child and parent.

0 = Not specified 1 = specified for parent or child 2 = specified for parent and child

1 Adapted from: The National Institute of Health Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. https://www.nhlbi.nih.gov/health-pro/guidelines/in-develop/cardiovascular-risk-reduction/tools/cohort

https://www.nhlbi.nih.gov/health-pro/guidelines/in-develop/cardiovascular-risk-reduction/tools/cohort

Table S3. Quality assessment of included studies

Article

Defined Sample

Represent-ative

Sample

Sample Size Attrition Missin

g Data

Valid Instrument

(Stress)

Valid Instrument

(Child Outcome)

Stress Measure

Child Outcome

Confou-nding

Variables

Demo. Info

Score (/15)

Alvik (2011) Y N N M Y Y N SR MR N P 5 Austin (2005) Y N N N Y Y Y SR MR Y P 6

Babineau (2015) Y Y N M Y Y Y SR MR Y C 9 Bolten (2012) N N N Y N Y Y SR MR Y C 7 Bolton (2013)1 Y N N Y N Y Y MM O N C 10 Bosquet Enlow (2017) Y Y N M N Y Y MM MR Y C 10 Brouwers (2001) N Y N M Y Y Y SR O Y C 10

Chittleborough (2011) Y Y N Y Y Y Y SR MM Y NS 10

Chong (2016) Y N N N Y Y Y SR MR Y P 6 Claridge (2015) Y N Y N Y Y Y SR MR Y C 8 Clark (2016) N N N Y Y Y Y SR MM Y NS 8 Coplan (2005) N N N M N Y Y SR MR N P 4 Crockenberg (2003) N N N M Y Y Y SR MR Y C 7 Cutrona (1986) N N N Y N Y Y SR MM N P 7 Davis (2012) Y N N Y N Y Y MM MR Y C 10 De Bruijn (2009) Y N N N N Y Y SR MM Y C 8

Della Vedova (2014) N Y N M Y Y Y SR MR Y P 7 Escallier (1995) N N N Y N Y N SR O Y C 7 Field (2009) N Y N Y N Y N SR O N C 7

Frost (1997) Y Y N Y Y Y Y SR MR Y P 11 Galbally (2015) Y N N M N Y Y SR MM Y C 9 Gerardin (2011) Y N N M N Y Y MM MM Y C 11 Godleski (2016) N N N Y N Y Y SR MM N C 8 Goyal (2009) N N N M Y Y N SR MM Y C 8 Grant (2010) Y N N N Y Y Y MM O Y C 10 Grazioli (2000) N N N Y N Y Y MM MR Y P 8 Gutteling (2005) Y N N N Y Y Y MM MM Y C 11 Guyon-Harris (2016) N N N Y Y Y Y SR MM Y P 9 Hanley (2013) Y Y N Y N Y Y MM O Y C 12 Hay (2008) N Y N M Y Y Y MM MM Y C 12 Hayatbakhsh (2012) Y N N M N Y Y SR MR Y C 7

Hosseini (2007) N N N M Y Y Y SR MR Y C 7 Huizink (2002) N N N M Y Y Y SR MM Y C 9 Husain (2012) Y N Y Y N Y Y MM O Y P 11 Isosavi (2017) Y N N Y Y Y Y MM MR Y C 11 Jones (2008) N N N M Y Y Y MM MR Y C 9 Korhonen (2012) Y N N N N Y Y SR MM Y C 8

Koutra (2013) Y Y N N Y Y Y SR MM Y C 10

Lin (2014) N N N N Y Y Y SR MM Y C 8 Lin (2017) Y Y N N Y Y Y MM MM Y C 12 Loutzenhiser (2015) N N Y Y Y Y Y SR MR N P 7 Luecken (2015) N N N M Y Y Y SR MM Y P 8 Macedo (2011) N N N N Y Y N MM MR N P 5 Mamott (1993) N N N Y N Y Y SR MR N P 5

McMahon (2013) N N N Y Y Y Y SR MR Y C 8

Moore (2016) N N N Y Y Y N MM O N C 9 Nolvi (2016) N N N Y N Y Y SR MR Y C 7 O'Connor (2003) Y Y N M Y Y Y SR MR Y P 8

Pacheco (2012) N N Y Y N Y Y SR O Y P 8 Peltola (2017) Y Y N Y Y Y Y SR MM Y P 11 Pihlakoski (2013) Y Y N M Y N Y SR MM Y P 9 Plant (2013) Y N N Y N Y Y O O Y C 10 Ramchandani (2010) Y N N N N N Y O MR Y C 6

Rautava (1993) Y Y N Y N N N SR MR Y P 6 Robinson (2008) Y Y N M Y N Y O MR Y C 9 Rode (2016) Y N N Y Y Y N SR MR Y P 7 Rothenberger (2011) Y N N Y N Y Y MM O N C 10 Rouse (2014) N N N M Y Y Y MM MR N C 8 Sandman (2015) N N N N N Y Y SR MR Y C 5 Sawada (2015) N Y N M Y Y N MM MR Y P 8 Sharp (2012) Y Y Y Y Y Y Y SR MM Y C 13 Sharp (2015) Y N N M N Y Y SR MR Y C 7 Stapleton (2012) N Y N N Y Y Y SR MR Y C 7 St James-Roberts (2005) N N N M Y Y Y SR MR N NS 4 Stroustrup (2016) Y Y N N N Y Y SR MM N P 7 Sullivan (2015) N N N Y Y Y Y SR MM Y C 10 Thomas (2017) Y Y N Y Y Y Y MM MM Y C 14 van den Berg (2009) Y Y N M Y Y N SR MR Y C 8

van den Heuvel1 (2015) N Y N N Y Y Y SR MR Y C 7 Warren (2004) N N N Y N Y Y SR MR Y C 7 Woolhouse (2016) Y Y N M Y Y Y SR MR Y P 8 Wurmser (2006) Y N N Y Y Y Y SR MR N C 8 Zhu (2014) Y N N Y N N Y SR MM Y C 9

Total (N = 73)

Criteria N (%) Criteria N (%) Criteria N (%)

Defined Sample Yes No

39 (53.42) 34 (46.58)

Missing Data Yes No

44 (60.27) 29 (39.73)

Subjective vs. Object Measure (Child) Multiple Methods

Objective Measure Maternal Report

24 (32.88) 11 (15.07) 38 (52.05)

Representativeness Yes No

24 (32.88) 49 (67.12)

Valid Instrument (Stress) Validated

Non-Validated

68 (93.15) 5 (6.85)

Confounding Variables Yes No

58 (79.45) 15 (20.55)

Adequate Sample Size

Yes No

5 (6.85) 68 (93.15)

Valid Instrument (Child Outcome)

Validated Non-Validated

63 (86.30) 10 (13.70)

Demographic Information Specified for Parent and Child

Specified for Parent or Child Not Specified

47 (64.38) 23 (31.51) 3 (4.11)

Participation/Attrition Acceptable

Marginally Acceptable Not Acceptable

32 (43.84) 25 (34.25) 16 (21.92)

Subjective vs. Objective Measure (Stress)

Multiple Methods Objective Measure

Self-report

19 (26.03) 3 (4.11) 51 (69.86)

1 This study was not included in analyses as the effect size was an outlier N = No; Y = Yes Demo.Info. = Demographic Information Attrition: Y = acceptable; M = marginally acceptable; N = unacceptable Stress Measure: SR = self-report; MM = mixed methods Child Outcome: MM = mixed methods; O = observation; MR = maternal report only Demographics: C = complete; P = partial; NS = Not-specific

Table S4 Results of Moderator Analyses for the Association between Maternal Prenatal Depression and Child Socio-Emotional Development Categorical Moderators k OR 95% CI Homogeneity

Q P-

value Syndromal-Level Symptoms No Yes

37 14

1.60** 2.26**

1.46-1.75 1.86-2.78

9.34 .01

Socio-Demographic Risk No Yes

40 10

1.66** 2.24**

1.51-1.84 1.73-2.91

4.37 .05

Type of Child Measure Temperament Behavior Problems Crying/Colic

27 20 3

1.78** 1.74**

1.98**

1.56-2.02 1.49-2.04 1.14-3.42

.19

.91

Method of Assessing Child Behavior Questionnaire Structured Measure Observation

40 5 5

1.77** 1.68**

1.86*

1.59-1.96 1.29-2.19 1.16-2.99

0.16

.92

Postnatal Depression Control No Yes

39 11

1.79** 1.68**

1.62-1.99 1.32-2.13

0.25

.62

Continuous Moderators

k

b

95% CI

Z-value

P-

value Pregnancy Time Point 48 .006 -.007-.019 0.90 .37 Maternal Age Child age

48 50

-.019 -.001

-.043-.005 -.002-.001

-1.56 -0.94

.12

.35 Percent of Males in Sample 50 -.013 -.041-.016 -0.88 .38 Study Quality 50 -.034 -.082-.015 -1.37 .17 k = sample size; b = estimate; 95% CI = 95% confidence intervals; *p < .01; **p<.001; *p<.01

Table S5 Results of Moderator Analyses for the Association between Maternal Prenatal Anxiety and Child Socio-Emotional Development Categorical Moderators k OR 95% CI Homogeneity

Q P-

value Socio-Demographic Risk No Yes

4

37

1.444** 1.64**

1.34-1.56 1.34-2.03

1.24 .27

Maternal Stress Measure a State Trait Life Events Mix of measures

14 12 3

12

1.64*** 1.48***

1.29 1.36***

1.41-1.90 1.28-1.70 0.96-1.74 1.22-1.52

4.34 .23

Type of Child Measure Temperament Behavior Problems Colic/Crying

23 9 7

1.51*** 1.55***

1.41***

1.37-1.68 1.37-1.77 1.41-1.72

0.68

.72

Method of Assessing Child Behavior Questionnaire Structured Measure Observation

30 5 4

1.49*** 1.27**

1.39

1.37-1.62 1.02-1.58 0.92-2.11

1.83

.40

Postnatal Depression Control No Yes

33 8

1.47** 1.49**

1.35-1.61 1.29-1.71

0.01

.91

Continuous Moderators

k

b

95% CI

Z-value

P-

value Pregnancy Time Point 40 .007 -.004-.017 1.30 .20 Maternal Age Child age

37 40

-.001 -.000

-.018-.018 -.002-.011

0.00 -0.57

.99

.57 Percent of Males in Sample 40 -.012 -.002-.025 1.67 .10 Study Quality 41 -.008 -.017-.002 -1.52 .13 k = sample size; b estimate; 95% CI = 95% confidence intervals; *p < .05; **p < .01; ***p<.001

a There were too few studies exclusively examining prenatal anxeity via prenatal cortisol to include these measures in this moderator analysis.

Figure S1. Funnel plot of the meta-analysis of included studies on prenatal depression and child socio-emotional development. Legend: The y-axis on the funnel plot represents the standard error, and the x-axis is the effect size. The white circles indicate studies that were included in the meta-analysis, and black circles indicate values to adjust for asymmetry in the funnel plot. The white diamond at the bottom of the funnel plot represents the observed mean effect size, and the black diamond represents the adjusted mean effect size.

Figure S2. Funnel plot of the meta-analysis of included studies on prenatal anxiety and child socio-emotional development. Legend: The y-axis on the funnel plot represents the standard error, and the x-axis is the effect size. The white circles indicate studies that were included in the meta-analysis, and black circles indicate values to adjust for asymmetry in the funnel plot. The white diamond at the bottom of the funnel plot represents the observed mean effect size, and the black diamond represents the adjusted mean effect size.

Supplemental References - All references included in Table 1 1. Alvik A, Torgersen AM, Aalen OO, Lindemann R. Binge alcohol exposure once a week

in early pregnancy predicts temperament and sleeping problems in the infant. Early Hum Dev. 2011;87(12):827-833.

2. Austin MP, Hadzi-Pavlovic D, Leader L, Saint K, Parker G. Maternal trait anxiety, depression and life event stress in pregnancy: relationships with infant temperament. Early Hum Dev. 2005;81(2):183-190.

3. Babineau V, Green CG, Jolicoeur-Martineau A, et al. Prenatal depression and 5-HTTLPR interact to predict dysregulation from 3 to 36 months--a differential susceptibility model. J Child Psychol Psychiatry. 2015;56(1):21-29.

4. Bolten MI, Fink NS, Stadler C. Maternal self-efficacy reduces the impact of prenatal stress on infant's crying behavior. J Pediatr. 2012;161(1):104-109.

5. Bolten M, Nast I, Skrundz M, Stadler C, Hellhammer DH, Meinlschmidt G. Prenatal programming of emotion regulation: neonatal reactivity as a differential susceptibility factor moderating the outcome of prenatal cortisol levels. J Psychosom Res. 2013;75(4):351-357.

6. Enlow MB, Devick KL, Brunst KJ, Lipton LR, Coull BA, Wright RJ. Maternal Lifetime Trauma Exposure, Prenatal Cortisol, and Infant Negative Affectivity. Infancy. 2017;22(4):492-513.

7. Brouwers EPM, van Baar AL, Pop VJM. Maternal anxiety during pregnancy and subsequent infant development. Infant Behav Dev. 2001;24(1):95-105.

8. Chittleborough CR, Lawlor DA, Lynch JW. Young maternal age and poor child development: predictive validity from a birth cohort. Pediatrics. 2011;127(6):e1436-1444.

9. Chong SC, Broekman BF, Qiu A, et al. Anxiety and Depression during Pregnancy and Temperament in Early Infancy: Findings from a Multi-Ethnic, Asian, Prospective Birth Cohort Study. Infant Ment Health J. 2016;37(5):584-598.

10. Claridge AM. Pregnancy intentions of first time mothers: Depressive symptoms, parenting stress, coparenting satisfaction, and child behavioral outcomes over the first three years: Department of Family and Child Sciences, Florida State University; 2014.

11. Clark CA, Espy KA, Wakschlag L. Developmental pathways from prenatal tobacco and stress exposure to behavioral disinhibition. Neurotoxicol Teratol. 2016;53:64-74.

12. Coplan RJ, O'Neil K, Arbeau KA. Maternal anxiety during and after pregnancy and infant temperament at three months of age. Journal of Prenatal & Perinatal Psychology & Health. 2005;19(3):199-215.

13. Crockenberg SC, Leerkes EM. Parental acceptance, postpartum depression, and maternal sensitivity: mediating and moderating processes. J Fam Psychol. 2003;17(1):80-93.

14. Cutrona CE, Troutman BR. Social support, infant temperament, and parenting self-efficacy: a mediational model of postpartum depression. Child Dev. 1986;57(6):1507-1518.

15. Davis EP, Sandman CA. Prenatal psychobiological predictors of anxiety risk in preadolescent children. Psychoneuroendocrinology. 2012;37(8):1224-1233.

16. de Bruijn AT, van Bakel HJ, van Baar AL. Sex differences in the relation between prenatal maternal emotional complaints and child outcome. Early Hum Dev. 2009;85(5):319-324.

17. Della Vedova AM. Maternal psychological state and infant’s temperament at three months. Journal of Reproductive and Infant Psychology. 2014;32(5):520-534.

18. Escallier LA. Prenatal predictors of infant colic: Maternal-fetal attachment, maternal state anxiety and maternal hope: School of Nursing, Adelphi University; 1995.

19. Field T, Diego M, Hernandez-Reif M, Ascencio A. Prenatal dysthymia versus major depression effects on early mother-infant interactions: a brief report. Infant Behav Dev. 2009;32(1):129-131.

20. Frost LA. Postpartum distress in fathers: Predicting depressive symtoms, anxiety, and anger at one month postpartum University of Wisconsin-Madison; 1996.

21. Galbally M, Lewis AJ, Buist A. Child developmental outcomes in preschool children following antidepressant exposure in pregnancy. Aust N Z J Psychiatry. 2015;49(7):642-650.

22. Gerardin P, Wendland J, Bodeau N, et al. Depression during pregnancy: is the developmental impact earlier in boys? A prospective case-control study. J Clin Psychiatry. 2011;72(3):378-387.

23. Godleski SA, Eiden RD, Schuetze P, Colder CR, Huestis MA. Tobacco exposure and maternal psychopathology: Impact on toddler problem behavior. Neurotoxicol Teratol. 2016;57:87-94.

24. Goyal D, Gay C, Lee K. Fragmented maternal sleep is more strongly correlated with depressive symptoms than infant temperament at three months postpartum. Arch Womens Ment Health. 2009;12(4):229-237.

25. Grant KA, McMahon C, Reilly N, Austin MP. Maternal sensitivity moderates the impact of prenatal anxiety disorder on infant responses to the still-face procedure. Infant Behav Dev. 2010;33(4):453-462.

26. Grazioli R, Terry DJ. The role of cognitive vulnerability and stress in the prediction of postpartum depressive symptomatology. Br J Clin Psychol. 2000;39 ( Pt 4):329-347.

27. Gutteling BM, de Weerth C, Willemsen-Swinkels SH, et al. The effects of prenatal stress on temperament and problem behavior of 27-month-old toddlers. Eur Child Adolesc Psychiatry. 2005;14(1):41-51.

28. Guyon-Harris K, Huth-Bocks A, Lauterbach D, Janisse H. Trajectories of maternal depressive symptoms across the birth of a child: associations with toddler emotional development. Arch Womens Ment Health. 2016;19(1):153-165.

29. Hanley GE, Brain U, Oberlander TF. Infant developmental outcomes following prenatal exposure to antidepressants, and maternal depressed mood and positive affect. Early Hum Dev. 2013;89(8):519-524.

30. Hay DF, Pawlby S, Waters CS, Sharp D. Antepartum and postpartum exposure to maternal depression: different effects on different adolescent outcomes. J Child Psychol Psychiatry. 2008;49(10):1079-1088.

31. Hayatbakhsh MR, O'Callaghan MJ, Bor W, Williams GM, Najman JM. Association of breastfeeding and adolescents' psychopathology: a large prospective study. Breastfeed Med. 2012;7(6):480-486.

32. Hosseini SM. Maternal anxiety during pregnancy: its relations to birth outcomes and to offspring depression during late childhood and adolescence Graduate School of Public Health, University of Pittsburgh; 2006.

33. Huizink AC, de Medina PG, Mulder EJ, Visser GH, Buitelaar JK. Psychological measures of prenatal stress as predictors of infant temperament. J Am Acad Child Adolesc Psychiatry. 2002;41(9):1078-1085.

34. Husain N, Cruickshank JK, Tomenson B, Khan S, Rahman A. Maternal depression and infant growth and development in British Pakistani women: a cohort study. BMJ Open. 2012;2(2):e000523.

35. Isosavi S, Diab SY, Kangaslampi S, et al. Maternal Trauma Affects Prenatal Mental Health and Infant Stress Regulation among Palestinian Dyads. Infant Ment Health J. 2017;38(5):617-633.

36. Jones SM. Maternal cortisol as a mediator of prenatal sterss and infant regulation development, Michigan State University; 2008.

37. Korhonen M, Luoma I, Salmelin R, Tamminen T. A longitudinal study of maternal prenatal, postnatal and concurrent depressive symptoms and adolescent well-being. J Affect Disord. 2012;136(3):680-692.

38. Koutra K, Chatzi L, Bagkeris M, Vassilaki M, Bitsios P, Kogevinas M. Antenatal and postnatal maternal mental health as determinants of infant neurodevelopment at 18 months of age in a mother-child cohort (Rhea Study) in Crete, Greece. Soc Psychiatry Psychiatr Epidemiol. 2013;48(8):1335-1345.

39. Lin B, Crnic KA, Luecken LJ, Gonzales NA. Maternal prenatal stress and infant regulatory capacity in Mexican Americans. Infant Behav Dev. 2014;37(4):571-582.

40. Lin Y, Xu J, Huang J, et al. Effects of prenatal and postnatal maternal emotional stress on toddlers' cognitive and temperamental development. J Affect Disord. 2017;207:9-17.

41. Loutzenhiser L, McAuslan P, Sharpe DP. The trajectory of maternal and paternal fatigue and factors associated with fatigue across the transition to parenthood. Clinical Psychologist. 2015;19:15-27.

42. Luecken LJ, MacKinnon DP, Jewell SL, Crnic KA, Gonzales NA. Effects of prenatal factors and temperament on infant cortisol regulation in low-income Mexican American families. Dev Psychobiol. 2015;57(8):961-973.

43. Macedo A, Marques M, Bos S, et al. Mother's personality and infant temperament. Infant Behav Dev. 2011;34(4):552-568.

44. Mamott BD. The transition to parenthood: The impact on quality of life Psychology, Illinois Institute of Technology; 1993.

45. McMahon CA, Boivin J, Gibson FL, et al. Pregnancy-specific anxiety, ART conception and infant temperament at 4 months post-partum. Hum Reprod. 2013;28(4):997-1005.

46. Moore GA, Quigley KM, Voegtline KM, DiPietro JA. Don't worry, be (moderately) happy: Mothers' anxiety and positivity during pregnancy independently predict lower mother-infant synchrony. Infant Behav Dev. 2016;42:60-68.

47. Nolvi S, Karlsson L, Bridgett DJ, et al. Maternal prenatal stress and infant emotional reactivity six months postpartum. J Affect Disord. 2016;199:163-170.

48. O'Connor TG, Heron J, Golding J, Glover V, Team AS. Maternal antenatal anxiety and behavioural/emotional problems in children: a test of a programming hypothesis. J Child Psychol Psychiatry. 2003;44(7):1025-1036.

49. Pacheco A, Figueiredo B. Mother's depression at childbirth does not contribute to the effects of antenatal depression on neonate's behavioral development. Infant Behav Dev. 2012;35(3):513-522.

50. Peltola MJ, Makela T, Paavonen EJ, et al. Respiratory sinus arrhythmia moderates the impact of maternal prenatal anxiety on infant negative affectivity. Dev Psychobiol. 2017;59(2):209-216.

51. Pihlakoski L, Sourander A, Aromaa M, et al. Do antenatal and postnatal parental psychological distress, and recognized need of help predict preadolescent's psychiatric symptoms? The Finnish Family Competence Cohort study. Child Psychiatry Hum Dev. 2013;44(2):305-319.

52. Plant DT, Barker ED, Waters CS, Pawlby S, Pariante CM. Intergenerational transmission of maltreatment and psychopathology: the role of antenatal depression. Psychol Med. 2013;43(3):519-528.

53. Ramchandani PG, Richter LM, Norris SA, Stein A. Maternal prenatal stress and later child behavioral problems in an urban South African setting. J Am Acad Child Adolesc Psychiatry. 2010;49(3):239-247.

54. Rautava P, Helenius H, Lehtonen L. Psychosocial predisposing factors for infantile colic. BMJ. 1993;307(6904):600-604.

55. Robinson M, Oddy WH, Li J, et al. Pre- and postnatal influences on preschool mental health: a large-scale cohort study. J Child Psychol Psychiatry. 2008;49(10):1118-1128.

56. Rode JL, Kiel EJ. The mediated effects of maternal depression and infant temperament on maternal role. Arch Womens Ment Health. 2016;19(1):133-140.

57. Rothenberger SE, Resch F, Doszpod N, Moehler E. Prenatal stress and infant affective reactivity at five months of age. Early Hum Dev. 2011;87(2):129-136.

58. Rouse MH, Goodman SH. Perinatal depression influences on infant negative affectivity: timing, severity, and co-morbid anxiety. Infant Behav Dev. 2014;37(4):739-751.

59. Sandman CA, Buss C, Head K, Davis EP. Fetal exposure to maternal depressive symptoms is associated with cortical thickness in late childhood. Biol Psychiatry. 2015;77(4):324-334.

60. Sawada N, Gagne FM, Seguin L, et al. Maternal prenatal felt security and infant health at birth interact to predict infant fussing and crying at 12 months postpartum. Health Psychol. 2015;34(8):811-819.

61. Sharp H, Pickles A, Meaney M, Marshall K, Tibu F, Hill J. Frequency of infant stroking reported by mothers moderates the effect of prenatal depression on infant behavioural and physiological outcomes. PLoS One. 2012;7(10):e45446.

62. Sharp H, Hill J, Hellier J, Pickles A. Maternal antenatal anxiety, postnatal stroking and emotional problems in children: outcomes predicted from pre- and postnatal programming hypotheses. Psychol Med. 2015;45(2):269-283.

63. Stapleton LR, Schetter CD, Westling E, et al. Perceived partner support in pregnancy predicts lower maternal and infant distress. J Fam Psychol. 2012;26(3):453-463.

64. St James-Roberts I, Conroy S. Do pregnancy and childbirth adversities predict infant crying and colic? Findings and recommendations. Neurosci Biobehav Rev. 2005;29(2):313-320.

65. Stroustrup A, Hsu HH, Svensson K, et al. Toddler temperament and prenatal exposure to lead and maternal depression. Environ Health. 2016;15(1):71.

66. Sullivan EL, Holton KF, Nousen EK, et al. Early identification of ADHD risk via infant temperament and emotion regulation: a pilot study. J Child Psychol Psychiatry. 2015;56(9):949-957.

67. Thomas JC, Letourneau N, Campbell TS, Tomfohr-Madsen L, Giesbrecht GF, Team APS. Developmental origins of infant emotion regulation: Mediation by temperamental negativity and moderation by maternal sensitivity. Dev Psychol. 2017;53(4):611-628.

68. van den Berg MP, van der Ende J, Crijnen AA, et al. Paternal depressive symptoms during pregnancy are related to excessive infant crying. Pediatrics. 2009;124(1):e96-103.

69. van den Heuvel MI, Johannes MA, Henrichs J, Van den Bergh BR. Maternal mindfulness during pregnancy and infant socio-emotional development and temperament: the mediating role of maternal anxiety. Early Hum Dev. 2015;91(2):103-108.

70. Warren LZ. Correlates and predictors of parenting and problem behaviors in young children of African American adolscent mothers: Psychology, University of Illinois at Urbana-Champaign; 2004.

71. Woolhouse H, Gartland D, Mensah F, Giallo R, Brown S. Maternal depression from pregnancy to 4 years postpartum and emotional/behavioural difficulties in children: results from a prospective pregnancy cohort study. Arch Womens Ment Health. 2016;19(1):141-151.

72. Wurmser H, Rieger M, Domogalla C, et al. Association between life stress during pregnancy and infant crying in the first six months postpartum: a prospective longitudinal study. Early Hum Dev. 2006;82(5):341-349.

73. Zhu P, Sun MS, Hao JH, et al. Does prenatal maternal stress impair cognitive development and alter temperament characteristics in toddlers with healthy birth outcomes? Dev Med Child Neurol. 2014;56(3):283-289.

a meta-analysis of maternal prenatal depression and ... · impacts on fetal neurodevelopment....

Documents